Can Crowdsourcing Improve Open Data?
What happens when open data is wrong? Can crowdsourcing improve it? Often, open data enthusiasts assume that the next step after the release of some government data is a smart phone app or cool visualisation. I’m more interested in collaborating on the data itself.
I’ve been working on a project called OpenEcoMaps and I’ve made use of open data releases, for which I’m very grateful. But the project is really focused on improving the data for London and making it useful for groups who don’t have access to smartphone developers with hip haircuts.
Take the Datastore entry on allotments, for example. The data was collected from boroughs by a London Assembly committee a few years back, and while it seems fairly comprehensive it only has midpoints rather than the shape of each allotment, and being a few years old it includes allotments that no longer exist. It also doesn’t include any community growing spaces, such as the thousands in the new Capital Growth network.
It turns out Capital Growth don’t have a very good dataset, either. They’ve got a great “wow factor” map but many locations are very vague and they don’t know the size of the growing spaces.
Councils don’t really know what’s out there either, though some such as Southwark (where I live) have commissioned organisations to map food projects for them.
I found a similar situation looking at renewable energy generators for a project local to me called Peckham Power. DECC have a map with barely anything on it; councils don’t keep a list of operating equipment from planning applications or their own estate; the GLA group have at least been slowly adding their renewables data to the Datastore following a request from the people at Peckham Power.
What obviously this calls for is open data collaboration. As it’s geodata I’m using OpenStreetMap to store the data – it’s a semi-mature technology in wide use, and it makes it reasonably easy for geeks to enter data and to pull it out for use in web apps, GIS software, etc.
OpenEcoMaps is basically a frontend, making the data in OpenStreetMap easy for people to use and contribute to. The idea is to make the data useful enough for a wide variety of organisations – public sector, charities, community groups, maybe even companies – to feel it’s worth collaborating on gathering and improving the data.
You can browse the maps, embed them on your web sites, use the KML files on a map you already have, and use a customised editor to contribute.
It’s not quite ready for use by your average local community group, but we’ve already got people using it in towns across the UK. It shows what you can do with open technology and data.
I’ve been meeting with the people behind London’s food strategy, Capital Growth, Southwark’s food strategy and various community food projects in my own neighbourhood to pilot OpenEcoMaps outside the map geek bubble.
Right now I really need help improving the code, so please spread the word. Over the next couple of years my hope is that this will spur growing interest amongst data hoarders in data collaboration rather than plain old data dumps.